Quiz 2.1-2.4

What is a simple random sample?Terms related?

• SRS: Each member of the population has an equal chance to be chosen for the sample; low bias

• To explain:

  • Label every individual in the population 1-N.

  • Generate x unique numbers using RandInt (1,N)

  • Select the individuals that correspond to these numbers chosen for the sample.

• Sampling without replacement: An individual can only be chosen once

  • Sampling with replacement: An individual can be chose twice.

General Survey Terms

• Population: The entire group we wish to study

• Census: Collecting info from an entire population

• Sample: A subset of the population that you collect data from.

• Sample Survey: Data collection from a sample

How to Sample Poorly:

• Convience Sample: Select individuals who are easiest to reach for the sample (high bias)

• Voluntary Response Sample: Indivduals chosen to be in a sample by responding to a general appeal

• Bias: When we systematically over or underestimate a value in the population.

Table of Random Digits

Use for random digits; if one value is 2 digits, all must be two; add zeroes as needed

Stratified Random Sample and what is true of good sampling?

• Stratified Random Sample: Split population into groups (strata) based on a variable that will affect results, pick a SRS from each, and combine

  • Low bias/variability

• To divide:

  • Split into groups based on ___.

  • Assign each person in group number 1 through n, where n is the total number of people in the population

  • Pick x random, unique numbers from each group, using RandInt (1,n) in the calculator

  • Etc.

• Good sampling: estimates of the population; low bias and low variability

Cluster Samples

• Cluster sample: divide the population into intact groups (clusters); randomly select clusters and survey every individual in the clusters and combine for the sample

  • Low bias/variability; easily accessible survey group

• Systematic Random Sample: Choose a random start point. Select every nth individual for the sample after this individual (N/n). Reduce bias.

Messing up, even when you’ve used a good sampling method…

• Undercoverage: portions of the population are left out from being a part of the sample (only asking 1 side of town, etc)

• Non-Response: Not everyone chosen for the sample responds

• Response Bias: pattern of incorrect responses

  • Wearing uniform